Brief negotiated interviews (BNI) are effective in linking alcohol abusers to treatment, but have not been adequately tested among crack/cocaine users. Using a randomized controlled trial, Project Link will test effectiveness of the BNI and an active referral process. 17,600 patients will be screened in the Urgent Care and Women's Health clinics at BMC, using DAST and frequency of use questions embedded in a Health Needs History. We expect 10% of patients with crack/cocaine and/or heroin use and DAST score>=3; they will be offered enrollment. All assessed at baseline by researchers blinded to experimental status, with biochemical analysis of hair, the Addiction Severity Instrument (ASI), Readiness to Change Ruler, and breathalizer analysis for alcohol. A control group of 800 will receive standard care (written advice/referral); intervention group of 800 will receive same plus the BNI/active referral. Follow up at 3 months will include structured interview and ASI. Follow up at 6 months will include same plus biochemical analysis